The Vavilov Framework V: Concluding Thoughts

In my first post on the Supreme Court of Canada’s decision in Vavilov I described it as a commendable effort to comprehensively address two issues, namely the selection of the standard of review and the application of the reasonableness standard, which have plagued Canadian administrative law for decades. Significantly, judges with divergent views on the rule of reviewing courts in administrative law cases formed a coalition coalescing around a new framework. In my analysis of the framework (Part I, Part II, Part III and Part IV), I identified some tensions and technical details which will have to be resolved or clarified in future cases, either by the Supreme Court or by lower courts. 

There is much to like in Vavilov, great chunks of which resemble the scholarly and judicial advice provided to the Supreme Court over the last decade. But will the majority reasons in Vavilov prove robust enough to provide a stable framework for Canadian administrative law?  In this concluding post I first note the scale of the changes effected in Vavilov, then identify some unusual consequences which will arise from the application of the Vavilov framework and describe how Vavilov essentially represents a pragmatic response before offering final thoughts on whether this pragmatic response will at long last bring clarity and predictability to Canadian administrative law.


There will certainly be significant changes: prior to Vavilov, judicial review principles applied to all appeal clauses. Some areas, admittedly, were only recently colonized by the marauding presumption of reasonableness review. Accordingly, Vavilov will represent a welcome return to the status quo ante in the municipal taxation regime considered in Edmonton East and other areas, such as expropriation, where the presumption was regarded as an invader

In other areas, Vavilov sweeps away decades of settled jurisprudence. Decisions of economic regulators, such as the federal CRTC and the provincial securities commissions, are typically subject to appeal clauses, as are the decisions of professional disciplinary tribunals. Of course, the courts had managed, in respect of some bodies (the Competition Tribunal springs to mind) to reason their way to correctness review notwithstanding the strong doctrinal currents which dragged them towards deference. But these are exceptions, not the rule and in general such entities have long been used to deference, even on questions of law. I dare say that this has not been especially controversial, given recognition that matters on which regulators have expertise can bleed into the interpretation of terms in their parent statutes. With expertise shunted to the margins, however, deference will no longer be the starting point in respect of these bodies.

It is, nonetheless, difficult to be categorical about the likely consequences of Vavilov for economic regulation and professional discipline. For although the appellate standard of review framework set out in Housen v Nikolaisen calls for correctness on questions of law, it calls for the deferential standard of palpable and overriding error on everything else, including mixed questions of law and fact. Much will depend, therefore, on the willingness of trial judges to categorize matters coming within the expertise of regulators as questions of law or as mixed questions. In that regard, it is worth mentioning that Housen is actually an unfamiliar standard for first-instance judges. Those who have come of (judicial) age post-New Brunswick Liquor may continue to incline towards deference. Deference might not be dead yet.

In terms of reasonableness review, the changes wrought by Vavilov are less significant though they may be consequential. In Vavilov the Supreme Court came down quite firmly on the side of hands-on reasonableness review (as exemplified by Delta Air Lines) rather than hands-off reasonableness review (as exemplified by Newfoundland Nurses), especially in its insistence that deference is to be founded in reasoned decisions and demonstrable expertise. 

In one sense, this is not a significant change: I would argue that, read properly, Newfoundland Nurses was not a blank cheque for administrative decision-makers and that the line of cases permitting ex post rationalizations of defective reasons was to be treated with caution. In another sense, Supreme Court confirmation that reasonableness review is to be “robust”, accompanied by equivocal guidance on jurisdictional error and statutory interpretation, may lead reviewing courts to more intensive scrutiny of administrative decisions. The clear requirement to start with the reasons for decision and work from there to a conclusion of unreasonableness  should stand in the way of overly aggressive review. But only time will tell.

The emphasis on remedial discretion is also new. Expect first-instance judges to give more consideration than heretofore on to whether and when to remit defective administrative decisions. As Brendan van Niejenhuis rightly observes, the law here will have to be built from the ground up. I would have liked the Supreme Court to provide more in the way of principled guidance to lower courts and to emphasize the systemic effects of exercises of remedial discretion. Lower courts may well provide what the Supreme Court did not. I hope they do. But again, only time will tell.

In some other areas the status quo has been maintained. Confusion continues to reign in respect of judicial review of delegated legislation. Correctness remains the standard on constitutional questions. I expect that there will be some pressure for Doré to be partly or wholly assimilated to the category of questions of central importance and/or the constitutional questions category but Doré is safe for now. As for procedural fairness, the Baker factors should continue to govern. Lastly, there was no attempt to change the appellate standard of review set out in Agraira. I expect there will be some pressure for change here. If palpable and overriding error is the standard for appellate intervention on regulatory determinations of fact and exercises of discretion, there will be pressure for similar light touch appellate review to be exercised in appeals of first-instance judicial review decisions.

Unusual consequences 

Applying the Vavilov framework leads to some strange results. Consider two. 

Where there is a statutory appeal, any issue of fact, discretion or mixed law and fact will be subject to the palpable and overriding error standard. On judicial review, by contrast, robust reasonableness review will be applied to any such issue. More deference would be due, in other words, in situations where an appeal has been provided for. Not only does this make little sense as a matter of principle, it will make little sense in practice to, say, a securities trader whose livelihood is put at risk by a regulatory decision — good luck explaining to her why her prospects on appeal are less than her prospects would have been on judicial review; references to “institutional design” are unlikely to be compelling.

Another unusual consequence is that some tribunals previously considered to be experts will be due less deference than decision-makers whose claim to expertise is much less compelling. The trilogy provides an example: the Registrar of Citizenship is entitled to more deference than the CRTC on questions of law. By any standard, that is an unusual consequence. Even the Competition Tribunal, with its cohort of judges, would get less deference in the Vavilov framework than the Registrar, or for that matter, a part-time adjudicator hearing appeals from roadside policing decisions in British Columbia. 

This is not to say that the Vavilov framework will not work. Perhaps some unusual consequences are a price worth paying for greater clarity. A framework which is predictably perverse may be more acceptable than one which is unpredictable. It is to say, though, that the Vavilov framework is one based on pragmatism, not principle. 

Pragmatism  vs Principle

Pragmatism was the name of the majority’s game in Vavilov. Whereas in Dunsmuir the standard of review was seen as a crucible in which the clash between the principles of the rule of law and democracy forged detailed doctrinal rules, in Vavilov principle is almost entirely absent. The Dunsmuir principles are retained (at para. 2) but they exercise little discernible influence over the Vavilov framework.

In terms of selecting the standard of review, institutional design is invoked as a master principle. But if it is a principle, it is a very thin one as the many unusual consequences described in the previous section testify.

Principle is largely absent from the majority’s articulation of a methodology for reasonableness review. Some principles are mentioned, admittedly (at paras. 74, 95, 127, 133). It is not obvious, however, how the core propositions (set out by the majority and accepted by the concurring judges) flow from these or any other principles. They read, rather, like a pragmatic attempt at shaping Supreme Court precedent into a workable approach to reasonableness review. As for the contextual factors enunciated as guides to the application of the reasonableness standard, it is again far from obvious how they relate to or can be said to follow from general principles.

As for remedial discretion, here some principles appear. But they are not the Dunsmuir principles, they are not the principles invoked in the majority’s articulation of reasonableness review and they bear no resemblance to the institutional design principle around which the selection of the standard of review is said to revolve.


Vavilov might be said, then, to represent the triumph of pragmatism over principle. But will the Vavilov framework nonetheless deliver the certainty and predictability the majority aims for? To answer this question, it is necessary to assess the issues, tensions and sources of pressure generated by the majority reasons. Here are my top ten:

  • Will the correctness categories, especially the ‘general questions of central importance to the legal system’ category, be narrowly construed?
  • Will correctness review be embraced where an appeal has been provided for, or will judges continue to defer to expert agencies, either by giving significant weight to administrative interpretations of law or by classifying them as questions of mixed law and fact calling for deferential review?
  • Will courts welcome the possibility, especially on appeals, of segmenting administrative decisions into questions calling for correctness review and other questions calling for deferential review?
  • Will the bar for judicial intervention on the reasonableness standard be the absence of “justification, transparency and intelligibility” or the presence of “serious shortcomings” in the administrative decision?
  • Will reviewing courts refuse to supplement administrative decisions said to be defective?
  • Will the “elements” of reasonableness review be treated as a box-ticking exercise or a guide to a holistic assessment of the reasonableness of a decision?
  • Will the majority’s contextual considerations be treated as a laundry list of factors permitting immediate judicial intervention or (again) a guide to a holistic assessment of reasonableness?
  • Will questions of statutory interpretation and questions which are arguably jurisdictional attract something akin to correctness review?
  • Will courts tend towards substitution of judgement and/or refusal to remit in cases where reasons for an administrative decision are sparse or non-existent?
  • Will the retrofitting of precedents to the Vavilov framework lead to more or less intrusive review?

Some of these are essentially technical issues which will lend themselves to pragmatic resolution. Most, however, are more ideologically charged issues on which a judge will be influenced by his or her own view of the legitimacy of administrative decision-making and which might therefore cause the Vavilov majority coalition to fracture and lower courts to diverge in their uses of the Vavilov framework.

Of course, these issues, tensions and sources of pressure might be addressed, resolved and removed in a pragmatic fashion. The consensus memorialized in the Vavilov framework might be welcomed to such an extent that judges put their principled disagreement about Canadian administrative law to one side. This is certainly a possibility, but it is impossible to be certain about what will happen next. The Vavilov framework represents a consensus on the Supreme Court and, probably, a consensus in Canada’s legal community. But the important question is whether the consensus will survive.

This content has been updated on December 29, 2019 at 21:12.